On modal identification of structures from earthquake response signals by a refined Frequency Domain Decomposition approach

Output-only frequency domain system identification via a refined Frequency Domain Decomposition (rFDD) algorithm is adopted towards the assessment of current modal properties of civil engineering buildings under strong ground motion base excitations. The aim is to investigate the true potentialities of such a modal dynamic identification approach within seismic engineering assessment and monitoring contexts. Indeed, by taking as input signals for the identification procedure structural response recordings arising from earthquake excitations, the typical assumption of stationary Gaussian white noise input laying at the theoretical bases of classical FDD techniques no longer hold. Then, the present non-parametric system identification technique is originally coupled to a Smoothed Wigner-Ville Distribution (S-WVD), resulting in an effective, self-contained time-frequency analysis framework. Synthetic signals generated from several shear-type frames with variable features, which are taken as benchmark structures are prior adopted as a necessary validation condition. In this context, use is made of several seismic records, including a complete set coming from the extended FEMA P695 earthquake database, containing 50 multi-component (NS, WE) earthquake records, 22 “Far-Field” and 28 “Near-Field”. Attempts with single and multiple input strong motions are also considered. Subsequently, real building earthquake response data have been effectively adopted. Heavy-damped structures are simultaneously considered at this stage, which appears a challenging scenario in the current context of FDD modal identification. Also, along with natural frequency and mode shape evaluation, the effective estimation of the modal damping ratios is specifically addressed, leading to dedicated computational procedures. Original strategies and arrangements are discussed, as attached to the specific processing of earthquake response signals and to the associated modal identification procedures. Results from the achieved estimates are statistically-compared to the target values, in order to prove the efficiency of the developed framework in providing prompt and accurate estimates of all current strong ground motion modal parameters. At this stage, such analysis tool may be employed for convenient, extensive applications in the realm of Earthquake Engineering, towards possible Structural Health Monitoring and damage detection purposes.